Paper Citation Record · LEDGER

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models

As of 22 August 2026, this Paper Citation Record lists 30 of 30 outbound references and 1 inbound Pith citation observation for arXiv:2605.20177.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2605.20177 v1

Coverage vector

measured 30 of 30 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-07-11T11:50:26.030339Z

measured 31 of 31 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-22T06:32:14.747728+00:00

measured 1 of 1 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-07-31T14:46:30.334667Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: cited_works

Reference resolution

30 of 30 outbound references displayed

verified exact21
verified fuzzy2
unresolved1
parse uncertain0
malformed identifier1
metadata mismatch5

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 58b2735c-2917-443e-abb3-75e0bff75d96 · outbound

This paper cites Nore- geo: Non-reasoning geometry benchmark.arXiv preprint arXiv:2601.10254.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Nore- geo: Non-reasoning geometry benchmark.arXiv preprint arXiv:2601.10254

Reference 1

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.522642Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4b90cb14-efeb-4715-a8d3-433eb05e6166 · outbound

This paper cites Qwen3-VL Technical Report.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Qwen3-VL Technical Report

Reference 2

Resolution: verified exact; local_arxiv, observed 2026-05-20T05:13:21.535514Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation ec0b5e64-b557-4071-bc35-f14648cee6bc · outbound

This paper cites ShareGPT4V: Improving Large Multi-Modal Models with Better Captions.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models ShareGPT4V: Improving Large Multi-Modal Models with Better Captions

Reference 3

Resolution: metadata mismatch; local_arxiv, observed 2026-05-20T05:13:21.508096Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 8445d69d-6a82-432a-bee2-4873c9088011 · outbound

This paper cites Video-R1: Reinforcing Video Reasoning in MLLMs.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Video-R1: Reinforcing Video Reasoning in MLLMs

Reference 4

Resolution: verified exact; local_arxiv, observed 2026-05-20T05:13:21.545362Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 5715ec37-505e-47c5-b169-dff5fe7e45ab · outbound

This paper cites G-LLaVA: Solving Geometric Problem with Multi-Modal Large Language Model.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models G-LLaVA: Solving Geometric Problem with Multi-Modal Large Language Model

Reference 5

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.504876Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation e07336f4-347c-40a0-9901-4679646b9a37 · outbound

This paper cites CogVLM2: Visual Language Models for Image and Video Understanding.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models CogVLM2: Visual Language Models for Image and Video Understanding

Reference 6

Resolution: metadata mismatch; local_arxiv, observed 2026-05-20T05:13:21.481162Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 51ea26a9-8f75-4e2d-bd1e-165529af6f56 · outbound

This paper cites Do Vision-Language Models Really Understand Visual Language?.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Do Vision-Language Models Really Understand Visual Language?

Reference 7

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.514602Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 3e7971cf-bbaf-4d2c-8221-235a3d504af6 · outbound

This paper cites Open-Reasoner-Zero: An Open Source Approach to Scaling Up Reinforcement Learning on the Base Model.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Open-Reasoner-Zero: An Open Source Approach to Scaling Up Reinforcement Learning on the Base Model

Reference 8

Resolution: verified exact; local_arxiv, observed 2026-05-20T05:13:21.561611Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation ec38e397-937d-4a55-b115-949d5acba2ab · outbound

This paper cites arXiv preprint arXiv:2508.02669 , year=.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models arXiv preprint arXiv:2508.02669 , year=

Reference 9

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.525956Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation d7aa1115-cc2c-4ed1-9598-d22cb6d7878d · outbound

This paper cites GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering

Reference 10

Resolution: metadata mismatch; local_arxiv, observed 2026-05-20T05:13:21.546757Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4a1f0191-67c1-401c-a645-e41a857a2b47 · outbound

This paper cites VisOnlyQA: Large Vision Language Models Still Struggle with Visual Perception of Geometric Information.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models VisOnlyQA: Large Vision Language Models Still Struggle with Visual Perception of Geometric Information

Reference 11

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.501063Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 9fa14ff1-e144-4db8-936a-0e635473e881 · outbound

This paper cites Mmr1: Enhancing multimodal reasoning with variance-aware sampling and open resources.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Mmr1: Enhancing multimodal reasoning with variance-aware sampling and open resources

Reference 12

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.560557Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation c5eaa9fa-2347-4767-9243-568217237c48 · outbound

This paper cites Multimodal ArXiv: A Dataset for Improving Scientific Comprehension of Large Vision-Language Models.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Multimodal ArXiv: A Dataset for Improving Scientific Comprehension of Large Vision-Language Models

Reference 13

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.507568Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 8ee5a468-1450-4a4b-baaf-9f19392b0e77 · outbound

This paper cites VisReason: A Large-Scale Dataset for Visual Chain-of-Thought Reasoning.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models VisReason: A Large-Scale Dataset for Visual Chain-of-Thought Reasoning

Reference 14

Resolution: verified exact; arxiv_id, observed 2026-07-02T01:17:25.615471Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 8161aa63-f862-49da-ae67-ed2dca173ef3 · outbound

This paper cites CLEVR-Math: A Dataset for Compositional Language, Visual and Mathematical Reasoning.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models CLEVR-Math: A Dataset for Compositional Language, Visual and Mathematical Reasoning

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.543836Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation ca1ce379-d2b8-419e-bb4b-08a71fff2d98 · outbound

This paper cites More Thinking, Less Seeing? Assessing Amplified Hallucination in Multimodal Reasoning Models.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models More Thinking, Less Seeing? Assessing Amplified Hallucination in Multimodal Reasoning Models

Reference 16

Resolution: metadata mismatch; arxiv_id, observed 2026-05-20T05:13:21.551928Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation fd501f9b-ba9e-4480-bdc0-61dba6a73518 · outbound

This paper cites Improved Baselines with Visual Instruction Tuning.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Improved Baselines with Visual Instruction Tuning

Reference 17

Resolution: verified exact; local_arxiv, observed 2026-05-20T05:13:21.557487Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 36e93f43-8a0a-4285-acb9-a2c247a2eb60 · outbound

This paper cites OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge

Reference 18

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.570421Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 1020f756-3f7b-4e10-911d-ebb048065042 · outbound

This paper cites ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning

Reference 19

Resolution: metadata mismatch; local_arxiv, observed 2026-05-20T05:13:21.573194Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 285a4026-1e00-4f33-a829-679f7d2deeb3 · outbound

This paper cites LMM-R1: Empowering 3B LMMs with Strong Reasoning Abilities Through Two-Stage Rule-Based RL.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models LMM-R1: Empowering 3B LMMs with Strong Reasoning Abilities Through Two-Stage Rule-Based RL

Reference 20

Resolution: verified exact; local_arxiv, observed 2026-05-20T05:13:21.567258Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation ae7bedce-d04a-4558-ab12-656c1ddcfc70 · outbound

This paper cites A-OKVQA: A Benchmark for Visual Question Answering using World Knowledge.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models A-OKVQA: A Benchmark for Visual Question Answering using World Knowledge

Reference 21

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.565954Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation ab60bffd-42be-4c82-a373-9b4b783f5001 · outbound

This paper cites Descriptive caption enhancement with visual specialists for multimodal perception.arXiv preprint arXiv:2412.14233.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Descriptive caption enhancement with visual specialists for multimodal perception.arXiv preprint arXiv:2412.14233

Reference 22

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.519154Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 04a38104-5c6b-48cf-8edf-7f6ac1384ef9 · outbound

This paper cites LlamaV-o1: Rethinking Step-by-step Visual Reasoning in LLMs.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models LlamaV-o1: Rethinking Step-by-step Visual Reasoning in LLMs

Reference 23

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.529260Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 11049ae2-ecda-41a2-9b5f-b6b5004bc4ba · outbound

This paper cites an unresolved cited work.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Unresolved cited work

Reference 24

Resolution: unresolved; raw_fallback, observed 2026-05-20T05:13:21.578814Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 6983eb48-fcc7-4ed4-9e77-1de6f3922bfd · outbound

This paper cites WeThink: Toward General-purpose Vision-Language Reasoning via Reinforcement Learning.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models WeThink: Toward General-purpose Vision-Language Reasoning via Reinforcement Learning

Reference 25

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.563530Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 8160789e-f389-48e2-ab36-3b6de0efeffc · outbound

This paper cites GThinker: Towards General Multimodal Reasoning via Cue-Guided Rethinking.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models GThinker: Towards General Multimodal Reasoning via Cue-Guided Rethinking

Reference 26

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.576609Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation d7a7893f-c13f-4f80-ba1d-8d33346d535b · outbound

This paper cites Improve Vision Language Model Chain-of-thought Reasoning.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Improve Vision Language Model Chain-of-thought Reasoning

Reference 27

Resolution: verified exact; arxiv_id, observed 2026-05-20T05:13:21.549579Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation b85715cc-147b-4f31-a7f8-6813f5d76688 · outbound

This paper cites Dynamath: A dynamic visual benchmark for evaluating mathematical reasoning robustness of vision language models.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Dynamath: A dynamic visual benchmark for evaluating mathematical reasoning robustness of vision language models

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-05-20T05:13:21.583182Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 44244442-f5a8-44d9-a6a3-b3e34868c57c · outbound

This paper cites Table 6.Key hyperparameters used in our Stage-3 training.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Table 6.Key hyperparameters used in our Stage-3 training

Reference 29

Resolution: verified fuzzy; raw_fallback, observed 2026-05-20T05:13:21.581098Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation dba7d2c7-074f-4993-aa57-2f2a59abf1ee · outbound

This paper cites an unresolved cited work.

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models Unresolved cited work

Reference 30

Resolution: malformed identifier; arxiv_id, observed 2026-05-20T05:13:21.555039Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Pith citing papers

Observation 7d07c8f7-0df1-49ad-83bf-b49a503c4082 · inbound

RP-OPSD: Resolution-Privileged On-Policy Self-Distillation for Multimodal Large Language Models cites this paper.

RP-OPSD: Resolution-Privileged On-Policy Self-Distillation for Multimodal Large Language Models From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models

Reference 67

Resolution: unresolved; no resolver link, observed 2026-07-31T14:46:30.334667Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.